Adam,
First - duh (
Adam Lally wrote:
This is actually expected. Some of our unit tests are testing error
conditions which cause the framework to log SEVERE messages.
) - the person who invents an automated, programmers' dope slap will
make a fortune
http://www.cartalk.com/content/read-on/2003/08.30-1.html
Second, using the -Dmaven.test.skip=true allows the build to complete.
Third, I have been "looking" more closely at class
MultiprocessingAnalysisEngine_implTest. I've run this test a number of
ways, all within Eclipse, and have seen different results. Here is a
summary:
1) running just this test as a junit test - it completes
2) running it under a code coverage tool - it completes
3) running it through the debugger - with multiple breakpoints - I
encountered a "hang up" on the third time through the loop for the code
(starting on line number 201) - sorry, I do not remember if it was the
creation or the start:
for (int i = 0; i < NUM_THREADS; i++) {
threads[i] = new ProcessThread(ae);
threads[i].start();
}
4) running it under the same code coverage tool - it hangs up.
I have not been able to reliably replicate the problem - I understand it
is a symptom of a race condition. I will continue trying to catch it
under the debugger and closer to "where" it is happening.
Bill
